DESCRIPTION: (Applicant's Description) The long-term goals of this proposal are to develop efficient molecular technologies which can be applied broadly to rapidly analyze clinical tumor specimens for differences in gene expression and identify at high resolution genomic losses/gains. Both the analysis of gene expression and genomic differences use hybridization of complex fluorescently labeled probes to micro-gridded DNA arrays. First, various conditions for the arraying of a set of 96 interalu products, generated from BACs that span chromosome 17 will be tested as hybridization targets for comparative genomic hybridization. These arrays will initially be validated by hybridization with DNA from somatic cell hybrids and tumor cell lines, Hybridization of DNA from neuroblastomas with known cytogenetic abnormalities on chromosome 17, as determined by CGH onto metaphase chromosomes, will serve as a final test of the ability of these arrays to detect ploidy difference in the context of the entire genome on clinical tumor samples. Conditions established in this proposal should be applicable to dense coverage of the whole genome for high resolution assessment of aneuploidies. Representational Difference Analysis will be optimized as a high throughput means to identify differentially expressed genes from primary tumor tissue. A system will be established for the rapid screening of differential expression of the RDA products on microarrays. A set of RDA products that represent differentially expressed genes from tumors with different behaviors (i.e., metastatic vs localized Ewing's sarcoma) will be arrayed and used to detect differentially expressed genes in primary tumor tissues. Methods developed in this proposal should be applicable to a wide spectrum of malignancies to better characterize and predict tumor behavior.